350 likes | 453 Views
Explore the user's journey in the ESG system, from search to data preview and analysis. Discover how the product subsystem facilitates data selection and visualization for a seamless experience. Delve into different experiments and modes to optimize data access and analysis. The story boards guide users through accessing, previewing, and analyzing data efficiently in the ESG Portal. Learn about the advanced features like zoom, time period selection, and shopping basket management for enhanced usability. Take a deep dive into the process of requesting visualizations and products, including options for online and offline data access. Improve your understanding of the ESG system and empower users to make informed decisions with this comprehensive exploration of the product subsystem.
E N D
Product-Generation in ESG:some explorations of the user experienceSteve Hankin – March, 2007
The user’s experience of ESG begins with the sign-on process. This story board picks up after that point … Continue Suppose the user elects to perform a search Search our holdings tofind data of interest? No thanks. I know whichdata I want to work with. Continue The user is offered a choice … ESG Story Poard – product subsystem
User constrains the search ESG will offer a search interface through which the user can discover runs of interest . (This Story Board will not attempt to explore the ESG search experience. RDBMS and OWL/RDF options are being examined in a separate discussion.) ESG Story Poard – product subsystem
Preview Preview reduced-resolution data Continue Obtain or analyze data data Experiment 2 Imagine a metadata summary of this experiment here … Next we’ll run through a scenario for “preview mode” access to the data. Question: Do the benefits of a preview mode justify the added complexity? The info icon leads to human- readable metadata User selects the run-variables of interest The result of a search is a list of “run-variables” (*) (*) A “run-variable” means a variable that is output by a model run. Even if the result of a search is simply a “model run”, that search result implicitly carries along with it a list of the output variables. ESG Story Poard – product subsystem
Experiment 4 ESG Portal – Preview reduced-resolution data Reduced resolution data for preview Experiment 2 – AAA Pre-industrial Naviga-tionbar of some kind … Experiment 4 – H&R winds ESG Story Poard – product subsystem
ESG Portal – Preview reduced-resolution data Reduced resolution data for preview Experiment 2 – AAA Pre-industrial Naviga-tionbar of some kind … Experiment 4 – H&R winds Temperature Salinity (no reduced resolution data available) Zonal velocity (no reduced resolution data available) Meridional velocity (no reduced resolution data available) User selects variable of interest A dimmed variable name indicates that no preview dataset is available ESG Story Poard – product subsystem
ESG Portal – Preview reduced-resolution data ESG Story Poard – product subsystem
Desktop: Matlab, IDL, IDV, Ferret, GrADS, … Rapid Access to Reduced Resolution Information Products • Custom subsets: • netCDF, • binary, • spreadsheet, • GIS layer, … ESG Story Poard – product subsystem
Preview Preview reduced-resolution data Continue Obtain or analyze data data But this time we click on “Continue” Lets return to our search results -- the list of “run-variables”. ESG Story Poard – product subsystem
Zoom User can zoom in on the time region of interest … We enter a tool to help visualize all of the available data selected from the search – both offline and on-line ESG Portal – Select time periods of interest Experiment 2 – AAA Pre-industrial Variables from Experiment 2 Naviga-tionbar of some kind … [+] Experiment 4 – H&R winds Variables from Experiment 4 time resolution [-] Temperature Preview Monthly 3-day 1-day Preview Salinity Monthly On-line 3-day Off-line 1-day Year: 2100 Year: 2000 ESG Story Poard – product subsystem
Zoom ESG Portal – Select time periods of interest Pick the variables and time resolutions he wants … Experiment 2 – AAA Pre-industrial Continue Variables from Experiment 2 Naviga-tionbar of some kind … [+] And continue to his shopping basket Experiment 4 – H&R winds Variables from Experiment 4 time resolution [-] Temperature Preview Monthly 3-day 1-day Preview Salinity Monthly On-line 3-day Off-line 1-day Year: 2070 Year: 2060 ESG Story Poard – product subsystem
ESG Portal – My Shopping Basket My Shopping Basket Or he can continue into the product generation system As we will nowdo … Naviga-tionbar of some kind … Offline data TapeToDisk [+] Experiment 4 Salinity 1-day 2060:2070 He can request migration of data from tape to disk (outside the scope of this Story Board) … FTP raw files Online data [-] Experiment 4 Temperature Preview 2000:2100 Continue Experiment 4 Temperature 3-day 2060:2070 He can FTP the raw files (also outside the scope of this Story Board) … Experiment 4 Temperature 1-day 2060:2070 Experiment 4 Salinity 3-day 2060:2070 Remove fromshopping basket ESG Story Poard – product subsystem
Experiment 2 Experiment 4 – H&R winds ESG Portal – Request visualizations and products All on-line data My Shopping Basket Naviga-tionbar of some kind … ESG Story Poard – product subsystem
Experiment 2 ESG Portal – Request visualizations and products All on-line data My Shopping Basket Naviga-tionbar of some kind … Experiment 4 – H&R winds Preview User selects data of interest Temperature 2000:2100 3-day resolution Temperature 2060:2070 Salinity 2060:2070 1-day resolution Temperature 2060:2070 ESG Story Poard – product subsystem
As before, user can select geometry, region, product and options … ESG Story Poard – product subsystem
Desktop: Matlab, IDL, IDV, Ferret, GrADS, … Continuing with a look at some of the subtler aspects of products … Information Products • Custom subset: • netCDF, • binary, • spreadsheet, • GIS layer, … ESG Story Poard – product subsystem
Analysiscapabilities Operation could be performed server-side or client-side … Define a new variable by applying simple transformations (averaging, variance, extrema, …)
Model Inter-comparison Compare fields When grids differ regridding is automated ESG Story Poard – product subsystem
Ensemble comparison Ensemble comparison is not just a viz style – it is a custom user interface “mode” ESG Story Poard – product subsystem
Demanding (large) products may be created in “batch” mode ESG Story Poard – product subsystem
Another example of server-side analysis On-the-fly regriddingfrom curvilinearto rectilinearcoordinates ESG Story Poard – product subsystem
Through OPeNDAP -- seamlessly segue from the browser to analysis in user’s favorite desktop tool ESG Story Poard – product subsystem
Server-side analysis available at the desktop, too The remote dataset is just an OPeNDAP URL http://server/my_output e.g. Vertical average of variable “TEMP” http://server/_expr_{my_output}{Tave=TEMP[Z=@AVE]} Define a new variable on-the-fly ESG Story Poard – product subsystem
Server-side analysis note: • “Delayed evaluation” • Calculations are on-demand, “surgical” • So, despite the huge size of the data, to the user the whole domain appears to have been transformed ESG Story Poard – product subsystem
e.g. Unix shell script access to products All ESG products (well, nearly) can also be made available through a service interface. In this example we show shell script access. Query available data sets: Query variables in data set “model_1”: Query space-time domain: Request a subset of data as a netCDF file: >lasls http://cpu/LAS >lasls http://cpu/LAS model_1 >lasls http://cpu/LAS model_1 sst >lasget -x 20:60 -y 20:60 -t 11-Dec-2000 -f netcdf http://cpu/LAS model_1 sst ESG Story Poard – product subsystem
The service interface also opens the door for special purpose UIsThis shows an example developed for K-12 students ESG Story Poard – product subsystem
Returning, finally, to the starting point…Here is a scenario for the advanced user, who has just signed-on and wants to return to his work from an earlier session. Continue Search our holdings tofind data of interest? No thanks. I know whichdata I want to work with. The returning user clicks here Continue ESG Story Poard – product subsystem
ESG Portal – Request visualizations and products open “all on-line data” All on-line data My Shopping Basket Naviga-tionbar of some kind … Experiment 4 – H&R winds Preview Temperature 2000:2100 3-day resolution Temperature 2060:2070 Salinity 2060:2070 1-day resolution Temperature 2060:2070 User may return to previous work on his shopping basket Or … ESG Story Poard – product subsystem
IPCC collection NCAR GFDL My Shopping Basket IRI COLA BADC UK Met Office ESG Portal – Request visualizations and products All on-line data Naviga-tionbar of some kind … IPCC collection ESG Story Poard – product subsystem
My Shopping Basket Experiment 6 GFDL My Shopping Basket IRI COLA BADC Experiment 7 Experiment 1 Experiment 2 Experiment 3 Experiment 4 UK Met Office Experiment 5 ESG Portal – Request visualizations and products All on-line data Naviga-tionbar of some kind … IPCC collection Preview Temperature 2000:2100 3-day resolution Temperature 2060:2070 Salinity 2060:2070 1-day resolution Temperature 2060:2070 ESG Story Poard – product subsystem
IRI Meridional Velocity Zonal Velocity COLA UK Met Office Temperature GFDL Experiment 4 Experiment 5 Experiment 6 Experiment 7 Salinity Experiment 3 ESG Portal – Request visualizations and products All on-line data Naviga-tionbar of some kind … IPCC collection Experiment 1 Preview 3-day resolution Temperature 2060:2070 Salinity 2060:2070 1-day resolution Temperature 2060:2070 ESG Story Poard – product subsystem
Preview Temperature Meridional Velocity Zonal Velocity Salinity Experiment 7 Experiment 6 Experiment 4 Experiment 3 Temperature Experiment 5 ESG Portal – Request visualizations and products All on-line data Naviga-tionbar of some kind … IPCC collection Experiment 1 Preview User selects an on-line aggregation of interest Preview Monthly 3-day 1-day Preview On-line 3-day resolution Off-line Temperature 2060:2070 Salinity 2060:2070 1-day resolution Temperature 2060:2070 ESG Story Poard – product subsystem
Once again, he enters the product-generation system. And as before … ESG Story Poard – product subsystem
Desktop: Matlab, IDL, IDV, Ferret, GrADS, … Information Products • Custom subset: • netCDF, • binary, • spreadsheet, • GIS layer, … ESG Story Poard – product subsystem
Discussion time … ESG Story Poard – product subsystem